Anthropic has completely suspended access to its newly launched Fable 5 and Mythos 5 artificial intelligence models. The shutdown occurred Friday night, just days after the models' release, following a direct directive from the US Commerce Department.
The government order subjects these specific models to strict export controls, restricting their use exclusively to the United States. Anthropic stated that an abrupt global disablement was the only immediate method to ensure full compliance with the federal mandate. Access to the company’s other AI models remains unaffected.
According to reports, the Trump administration acted on concerns regarding a reported jailbreak vulnerability. This exploit allegedly bypasses classifier-based safeguards designed to block prompts related to cybersecurity, chemistry, and biology. Administration officials requested a pause to allow the national security apparatus time to harden defenses against such threats, a process expected to take several weeks.
Anthropic contends that the government provided only verbal evidence of a narrow, non-universal jailbreak. The company notes the exploit involves reviewing codebases for software flaws, describing the identified vulnerabilities as minor and simple. Anthropic further argued that comparable capabilities exist in other publicly available models, including GPT-5.5.